Countersunk angles: 82°, 90°, and the hole

It looks like a choice, and it is drawn like one

Every countersunk screw catalogue lists both 82° and 90°. They sit next to each other like two finishes or two head sizes, so treating the angle as a preference is exactly what the presentation invites.

And most of the time nothing goes wrong, because a whole assembly usually comes from one standard and the angles match without anybody thinking about it. The problem only shows up when a drawing quotes one standard and the screw comes from another.

So pick 90° — except it was never yours to pick

The common misunderstanding is treating the angle as a preference. It follows the standard the drawing cites:

What a few degrees actually does

Treat the head and the countersink as two coaxial cones. Once the included angles differ, they can only meet on one circle. Which circle depends on whose cone is steeper:

• Screw angle smaller than hole angle (an 82° screw in a 90° hole): the head cone is steeper than the hole, so contact lands on the top rim and the head stands proud. Flush is not available.
• Screw angle larger than hole angle (a 90° screw in an 82° hole): the head cone is flatter, so contact drops to the small end at the bottom of the countersink, loading the thinnest part of the hole.

Wherever it lands, the consequences are the same set:

• Face contact becomes line contact. Local stress rises; thin sheet can dish, or the edge of the hole rolls over
• The head face is not parallel to the surface, so “flush” is only flush to the eye
• The friction radius changes, and with it the torque-to-clamp relationship. If the joint is torque controlled, those parameters were established on a correct pairing — the same argument appears in the stack-up article, and it does not survive a changed pairing.

And there is a second mechanism, which does not need anybody to have made a mistake at all. The hole and the head are tolerated in different documents, and their ranges are not centred on each other:

Which is why the angle stays in every permitted callout format, in both standards systems: two callout systems. And if the drawing is going to a Taiwanese supplier, note that the head has four names in Chinese and the standards use a different one from the trade.

Where 100°, 110° and 120° come from

It is a geometric trade, not vendors inventing variants. Opening the included angle buys a larger head diameter for the same head height, and a shallower countersink. When thin sheet has to keep bearing area while the head gets flatter, that is the direction with something to give.

Cabinet work is where we have actually seen demand move that way: thin panels, flush required, and a head that must not be small enough to pull through. That is what has come across our own desk — not a classification from a standard. 82° and 90° have standards behind them, 100° is common in aerospace series, and beyond that you are generally working to a drawing.

Thinner and smaller heads leave less to work with

The other half of the demand is the head itself. Three things move together:

• Bearing area falls. The same clamp load presses on a smaller annulus, so pull-through risk in thin sheet rises
• The countersink can break through the sheet. If head height exceeds sheet thickness the countersink cuts through and leaves a knife edge — which neither survives handling nor provides a bearing face
• Recess depth gets squeezed. A thinner head cannot carry a deep cross recess or hex socket, so driver engagement is shallower and cam-out becomes more likely

Flushness and angle are two requirements, not one

Most arguments about a countersink are really two requirements wearing one name. Installation specifications tend to control the installed height. Hole standards control the geometry. Neither one implies the other.

NASA PRC-9007C §3.0 is the clearest example of the first: it requires countersink diameter and depth to be specified such that installed fasteners meet a standard maximum protrusion of 0.010 in, unless the drawing gives an alternate tolerance. That is a statement about where the top of the head ends up — not about the cone.

The reverse case — angle within tolerance and the head still standing proud — has plenty of ordinary mechanisms, and none of them is an angle problem:

• Countersink major diameter undersized, or insufficient axial depth
• Screw head at its maximum actual diameter, or head height at its limit
• The head-to-shank fillet fouling the pilot hole edge — NASA-STD-5020A requirement TFSR 22 specifically requires clearance for it, via a chamfered hole or a countersunk washer
• Burrs, coating build-up, debris or chatter ridges
• Countersink and pilot hole eccentric, or their axes not parallel
• An edge break or secondary chamfer moving the surface-intersection datum
• Plate too thin to take the required cone without leaving a knife edge

Which is why “it will not sit flush” is not a diagnosis. The useful first question is whether the drawing is asking for a height or a geometry, because the two are inspected differently and can fail independently. The fillet clearance case is in the fillet under the head.

Four things to put on the drawing

• The head standard and its angle. ISO 10642 (90°) or ASME B18.6.3 (82°), not just “countersunk”
• The countersink angle and top diameter. The hole has standards of its own — ISO 15065 and DIN 74, with head profiles in ISO 7721. Published countersunk hole dimensions disagree with each other more often than people expect, so it matters which chart the number came from
• Flushness acceptance. How much proud and how much below, and measured how
• Sheet thickness against head height. Whether a knife edge is created is arithmetic you can do in advance

Five questions before you sample

• What angle is your hole? Punched, countersunk with a tool, or machined — the deviations differ
• How is flushness judged? By eye, by gauge, or by measuring protrusion
• How thick is the sheet, and does the head height fit?
• Has plating been accounted for? Coating on the cone moves where the head seats (the allowance problem is sharper at small sizes)
• Is this angle a standard item or made to drawing? The tooling and the lead time are not the same

Ask us the same question

What we can confirm: below M6 the wire, tooling, thread rolling and inspection are in house, forming, inspection and packing run in one flow, and have since 1994.

This page covers step 3, the head. The whole order — substrate, thread, head, drive, finish, documentation — and why doing it out of order is rework rather than a tweak, is in specifying a screw.

References

• Countersink — Wikipedia

Standards referred to above, listed by number for verification:

• ASME B18.6.3　Machine Screws, Tapping Screws, and Metallic Drive Screws (inch flat head, 82°; the same standard also carries 100° heads)
• NASA PRC-9007C　§3.0 — countersink diameter and depth specified for a standard maximum installed protrusion of 0.010 in
• NASA-STD-5020A　requirement TFSR 22 — clearance for the head-to-shank fillet, via a chamfered hole or a countersunk washer
• ISO 7046 / ISO 10642　metric countersunk head screws (90°)
• ISO 7721　countersunk flat head screws — head configuration
• ISO 15065 / DIN 74　countersinks for countersunk head screws

This page describes the geometry and what to specify. It is not a set of design values for any particular product. Angles, tolerances and acceptance criteria are governed by the standards themselves and by your drawing.